The HLA class I alleles B57 and B27 are associated with control of HIV. Unfortunately, most individuals do not express these protective alleles, and are thus unable to control viral replication and slow disease progression. If we could understand the mechanisms of this control, might we be able to instill the protective effects of these alleles to more common HLA alleles through vaccination? Growing data supports that the immune control elicited by HLA-B57 and -B27 is mediated in part by their dominant targeting of highly conserved regions of HIV by CD8+ T cell responses, suggesting that viral escape from these responses comes at a substantial cost to viral replication. Recent studies now illustrate that CTL escape mutations selected by these protective HLA alleles do substantially impair viral replication, while in contrast escape mutations selected by dominant CD8 responses restricted by other common HLA alleles do not impact viral replication. Therefore, the front-line CD8 responses elicited by most individuals are rapidly evaded with little to no cost to the virus. Many subdominant CD8 responses to common HLA alleles do, however, target highly conserved regions throughout HIV in which escape mutations might impair viral replication. Unfortunately, these subdominant responses don't arise during acute infection due to dominant CD8 responses that 'mask'the expansion of subdominant responses. We hypothesize that if specific subdominant responses were not masked that these responses might mediate early control of HIV. To test this hypothesis we will identify subdominant CD8 responses that select for uniquely deleterious CTL escape mutations in HIV and test their ability to enhance control of HIV in the BLT humanized mouse model. As a proof of concept, we will focus on four common HLA alleles (A02, A03, B07, B08) and will infect BLT mice with HIV strains in which the dominant CD8 epitopes have been 'knocked out'to determine whether refocusing CD8 responses against more critical regions of HIV results in more effective control. We will address the following specific aims: Aim 1: Identify subdominant CD8+ T cell responses for common HLA alleles that select for highly deleterious CTL escape mutations using in vitro replication assays, ultra-deep sequencing, and detection of variant-specific CTL responses. Aim 2: Determine whether HIV-specific CD8+ T cell immunodominance patterns and viral escape pathways are similar between humans and BLT humanized mice expressing common HLA alleles. Aim 3: Determine whether refocusing early CD8 responses against critical subdominant CD8 epitopes in humanized mice can enable early control of HIV in the setting of common HLA alleles. This study will identify subdominant CD8+ T cell responses capable of selecting for highly deleterious escape mutations in HIV that substantially impair viral replication. The potency of these responses will be tested in a new humanized mouse model to identify immune responses critical to the design of an effective HIV vaccine.